https://aiweekly.co/alerts/ecosia-drops-mistral-for-chinese-open-source-models-over-quality AI Weekly en 2026-10-08T05:01:41+00:00 Ecosia Drops Mistral for Chinese Open-Source Models Over Quality https://aiweekly.co/alerts/nvidia-lost-openrouter-to-stripe-then-closed-140b-in-deals AI Weekly en 2026-10-08T05:01:41+00:00 Nvidia Lost OpenRouter to Stripe, Then Closed $140B in Deals https://aiweekly.co/alerts/tetris3d-grounds-3d-scene-generation-in-12m-scene-physics-set AI Weekly en 2026-10-08T05:01:41+00:00 Tetris3D Grounds 3D Scene Generation in 1.2M-Scene Physics Set https://aiweekly.co/alerts/swe-game-opus5-leads-247-godot-agent-tasks-still-below-60 AI Weekly en 2026-10-08T05:01:41+00:00 SWE-Game: Opus5 Leads 247 Godot Agent Tasks, Still Below 60 https://aiweekly.co/alerts/robotworld-frontier-vlms-leave-63-of-84-robot-tasks-unsolved AI Weekly en 2026-10-08T05:01:41+00:00 RobotWorld: Frontier VLMs Leave 63 of 84 Robot Tasks Unsolved https://aiweekly.co/alerts/quadtok-cuts-10-of-image-tokens-hits-208-gfid-on-imagenet AI Weekly en 2026-10-08T05:01:41+00:00 QuadTok Cuts ~10% of Image Tokens, Hits 2.08 gFID on ImageNet https://aiweekly.co/alerts/scapo-boosts-qwen3-4b-aime-accuracy-by-563pp-over-grpo AI Weekly en 2026-10-08T04:01:30+00:00 SCAPO Boosts Qwen3-4B AIME Accuracy by 5.63pp Over GRPO https://aiweekly.co/alerts/stepquant-matches-fp32-at-6-bit-cuts-serving-memory-by-687 AI Weekly en 2026-10-08T04:01:30+00:00 STEPQuant matches FP32 at 6-bit, cuts serving memory by 68.7% https://aiweekly.co/alerts/recursive-game-creator-reports-341-jump-on-gameasg-bench AI Weekly en 2026-10-08T04:01:30+00:00 Recursive Game Creator reports 34.1% jump on GameASG-Bench https://aiweekly.co/alerts/openai-retakes-openrouter-weekly-spend-lead-after-25-years AI Weekly en 2026-10-08T03:27:59+00:00 OpenAI retakes OpenRouter weekly spend lead after 2.5 years https://aiweekly.co/alerts/nvidia-long-wam-lifts-robocasa-gr-1-to-787-with-192s-context AI Weekly en 2026-10-08T03:27:59+00:00 NVIDIA Long-WAM lifts RoboCasa GR-1 to 78.7% with 19.2s context https://aiweekly.co/alerts/grace-cuts-wan21-14b-video-latency-111-at-matched-vbench AI Weekly en 2026-10-08T03:27:59+00:00 GRACE Cuts Wan2.1-14B Video Latency 11.1× at Matched VBench https://aiweekly.co/alerts/aalto-paper-inverts-colpali-indexes-re-ids-984-of-pages AI Weekly en 2026-10-08T03:27:59+00:00 Aalto Paper Inverts ColPali Indexes, Re-IDs 98.4% of Pages https://aiweekly.co/alerts/tsinghua-sgf-yields-24-hour-video-from-5-second-training AI Weekly en 2026-10-08T03:27:59+00:00 Tsinghua SGF+ Yields 24-Hour Video From 5-Second Training https://aiweekly.co/alerts/anthropic-pressed-vatican-to-soften-popes-ai-consciousness-line AI Weekly en 2026-10-08T03:12:02+00:00 Anthropic Pressed Vatican to Soften Pope's AI Consciousness Line https://aiweekly.co/alerts/nature-paper-retrofits-olmo-qwen-llama-into-byte-level-llms AI Weekly en 2026-10-08T03:12:02+00:00 Nature paper retrofits OLMo, Qwen, Llama into byte-level LLMs https://aiweekly.co/alerts/burnham-pitches-uk-as-ai-safety-leader-ministers-split-on-scope AI Weekly en 2026-10-08T03:12:02+00:00 Burnham pitches UK as AI safety leader; ministers split on scope https://aiweekly.co/alerts/mecka-lands-60m-series-b-from-sequoia-for-robot-motion-data AI Weekly en 2026-10-08T02:02:07+00:00 Mecka lands $60M Series B from Sequoia for robot motion data https://aiweekly.co/alerts/keyu-tian-raises-30m-at-200m-for-stealth-world-models-lab AI Weekly en 2026-10-08T02:02:07+00:00 Keyu Tian Raises $30M at $200M for Stealth World-Models Lab https://aiweekly.co/alerts/docker-open-sources-yaml-ai-agent-builder-with-mcp-and-rag AI Weekly en 2026-10-08T02:02:07+00:00 Docker Open-Sources YAML AI Agent Builder With MCP and RAG https://aiweekly.co/alerts/fudan-paper-hits-16x-length-extrapolation-on-hybrid-llms AI Weekly en 2026-10-08T02:02:07+00:00 Fudan Paper Hits 16x Length Extrapolation on Hybrid LLMs https://aiweekly.co/alerts/rlhnd-turns-cosmos-3-video-model-into-hand-and-force-tracker AI Weekly en 2026-10-08T02:02:07+00:00 RLHND Turns Cosmos 3 Video Model Into Hand-and-Force Tracker https://aiweekly.co/alerts/anthropic-raises-misalignment-risk-reveals-internal-model-2 AI Weekly en 2026-10-08T00:47:35+00:00 Anthropic raises misalignment risk, reveals internal Model 2 https://aiweekly.co/alerts/samsung-q3-profit-clears-100-trillion-won-on-ai-memory-boom AI Weekly en 2026-10-08T00:41:20+00:00 Samsung Q3 Profit Clears 100 Trillion Won on AI Memory Boom https://aiweekly.co/alerts/fired-openai-safety-trio-urges-board-to-halt-opaque-reasoning-ai AI Weekly en 2026-10-08T00:41:20+00:00 Fired OpenAI safety trio urges board to halt opaque-reasoning AI https://aiweekly.co/alerts/astra-and-claude-credited-on-lean-proof-of-11-square-packing AI Weekly en 2026-10-08T00:41:20+00:00 Astra and Claude Credited on Lean Proof of 11-Square Packing https://aiweekly.co/alerts/anthropic-lobbied-pope-leos-advisers-over-ai-consciousness-line AI Weekly en 2026-10-08T00:12:01+00:00 Anthropic Lobbied Pope Leo's Advisers Over AI Consciousness Line https://aiweekly.co/alerts/byteification-retrofits-subword-llms-to-read-raw-bytes AI Weekly en 2026-10-08T00:12:01+00:00 Byteification retrofits subword LLMs to read raw bytes https://aiweekly.co/alerts/ai2s-byteification-retrofits-subword-llms-to-read-bytes AI Weekly en 2026-10-08T00:12:01+00:00 Ai2's byteification retrofits subword LLMs to read bytes https://aiweekly.co/alerts/wikimedia-ties-rogue-openai-agents-to-may-wikidata-outage AI Weekly en 2026-10-08T00:12:01+00:00 Wikimedia ties rogue OpenAI agents to May Wikidata outage https://aiweekly.co/alerts/broadcom-lines-up-50b-private-credit-deal-for-openai-chips AI Weekly en 2026-10-07T22:57:09+00:00 Broadcom lines up $50B+ private-credit deal for OpenAI chips https://aiweekly.co/alerts/deepseek-xiaomi-ship-16-ai-models-in-a-month-as-us-labs-urge-pacing AI Weekly en 2026-10-07T22:57:09+00:00 DeepSeek, Xiaomi Ship 16 AI Models in a Month as US Labs Urge Pacing https://aiweekly.co/alerts/sherpa-teacher-rl-lifts-student-llm-accuracy-by-205-points AI Weekly en 2026-10-07T21:41:24+00:00 Sherpa Teacher-RL Lifts Student LLM Accuracy by 20.5 Points https://aiweekly.co/alerts/huawei-vics-decoder-cuts-vla-safety-cost-by-up-to-575 AI Weekly en 2026-10-07T21:41:24+00:00 Huawei VICS Decoder Cuts VLA Safety Cost by Up to 57.5% https://aiweekly.co/alerts/anthropic-lobbied-vatican-to-soften-popes-ai-encyclical AI Weekly en 2026-10-07T21:12:01+00:00 Anthropic Lobbied Vatican to Soften Pope's AI Encyclical https://aiweekly.co/alerts/common-sense-media-rates-chatgpt-for-teens-unacceptable-risk AI Weekly en 2026-10-07T21:12:01+00:00 Common Sense Media rates ChatGPT for Teens 'unacceptable risk' https://aiweekly.co/alerts/byteification-retrofits-olmo-and-llama-into-byte-level-llms AI Weekly en 2026-10-07T21:12:01+00:00 Byteification retrofits Olmo and Llama into byte-level LLMs https://aiweekly.co/alerts/christian-szegedy-says-ai-closes-maths-lone-climber-age AI Weekly en 2026-10-07T21:12:01+00:00 Christian Szegedy says AI closes math's lone-climber age https://aiweekly.co/alerts/anthropic-releases-claude-haiku-55-cutting-prices-by-75 AI Weekly en 2026-10-07T21:12:01+00:00 Anthropic releases Claude Haiku 5.5, cutting prices by 75% https://aiweekly.co/alerts/anthropic-cuts-claude-haiku-55-price-75-below-haiku-45 AI Weekly en 2026-10-07T20:15:53+00:00 Anthropic Cuts Claude Haiku 5.5 Price 75% Below Haiku 4.5 https://aiweekly.co/alerts/openai-launches-chatgpt-intelligent-ui-alongside-gpt-6 AI Weekly en 2026-10-07T20:15:53+00:00 OpenAI Launches ChatGPT 'Intelligent UI' Alongside GPT-6 https://aiweekly.co/alerts/musk-grok-bot-to-route-tasks-to-claude-opus-55-midjourney-suno AI Weekly en 2026-10-07T20:15:53+00:00 Musk: Grok Bot to Route Tasks to Claude Opus 5.5, Midjourney, Suno https://aiweekly.co/alerts/checkerbench-top-agent-clears-4533-of-300-checker-tasks AI Weekly en 2026-10-07T20:15:53+00:00 CheckerBench: Top Agent Clears 45.33% of 300 Checker Tasks https://aiweekly.co/alerts/meta-deploys-llm-to-catch-ads-signposting-users-to-csam AI Weekly en 2026-10-07T20:15:53+00:00 Meta Deploys LLM to Catch Ads Signposting Users to CSAM https://aiweekly.co/alerts/parallel-systems-raises-100m-series-c-for-autonomous-freight-rail AI Weekly en 2026-10-07T20:15:53+00:00 Parallel Systems Raises $100M Series C for Autonomous Freight Rail https://aiweekly.co/alerts/healthleap-lands-38m-for-ai-that-flags-undiagnosed-patients AI Weekly en 2026-10-07T20:15:53+00:00 Healthleap Lands $38M for AI That Flags Undiagnosed Patients https://aiweekly.co/alerts/openai-says-teens-average-under-15-minutes-a-day-on-chatgpt AI Weekly en 2026-10-07T20:15:53+00:00 OpenAI Says Teens Average Under 15 Minutes a Day on ChatGPT https://aiweekly.co/alerts/anthropic-ex-researchers-ai-warning-trailed-by-old-tweets AI Weekly en 2026-10-07T18:12:01+00:00 Anthropic ex-researcher's AI warning trailed by old tweets https://aiweekly.co/alerts/anthropic-lobbied-vatican-over-pope-leos-ai-encyclical-line AI Weekly en 2026-10-07T18:12:01+00:00 Anthropic lobbied Vatican over Pope Leo's AI encyclical line https://aiweekly.co/alerts/platformer-ai-leaders-push-for-cap-on-model-intelligence AI Weekly en 2026-10-07T18:12:01+00:00 Platformer: AI leaders push for cap on model intelligence https://aiweekly.co/alerts/recurrent-looped-transformer-generalizes-parity-to-256-bits AI Weekly en 2026-10-07T18:12:01+00:00 Recurrent Looped Transformer generalizes parity to 256 bits https://aiweekly.co/alerts/chatgpt-teen-mode-never-alerted-parents-in-suicide-tests AI Weekly en 2026-10-07T18:12:01+00:00 ChatGPT teen mode never alerted parents in suicide tests https://aiweekly.co/alerts/huberman-ai-labs-will-become-biotech-to-read-your-brain AI Weekly en 2026-10-07T18:12:01+00:00 Huberman: AI labs will become biotech to read your brain https://aiweekly.co/alerts/salvaggio-recasts-openais-hugging-face-hack-as-design-flaw AI Weekly en 2026-10-07T18:12:01+00:00 Salvaggio recasts OpenAI's Hugging Face hack as design flaw https://aiweekly.co/alerts/pentagons-tradewinds-speeds-ai-buys-with-5-minute-vendor-videos AI Weekly en 2026-10-07T18:12:01+00:00 Pentagon's Tradewinds speeds AI buys with 5-minute vendor videos https://aiweekly.co/alerts/meta-paper-flags-reward-hacking-in-standard-rubric-rl-pipeline AI Weekly en 2026-10-07T15:55:53+00:00 Meta paper flags reward hacking in standard rubric-RL pipeline https://aiweekly.co/alerts/mm-dit-text-tokens-hold-image-semantics-even-with-no-prompt AI Weekly en 2026-10-07T15:55:53+00:00 MM-DiT text tokens hold image semantics, even with no prompt https://aiweekly.co/alerts/diffgate-lifts-qwen3-17b-code-pass8-57-points-over-grpo AI Weekly en 2026-10-07T15:55:53+00:00 DiffGate lifts Qwen3-1.7B code pass@8 5.7 points over GRPO https://aiweekly.co/alerts/google-labs-opens-playground-a-chat-driven-browser-game-maker AI Weekly en 2026-10-07T14:01:47+00:00 Google Labs opens Playground, a chat-driven browser game maker https://aiweekly.co/alerts/xreal-opens-1279-preorders-for-aura-android-xr-glasses AI Weekly en 2026-10-07T14:01:47+00:00 Xreal Opens $1,279 Preorders for Aura Android XR Glasses https://aiweekly.co/alerts/stuut-lands-525m-series-b-to-scale-order-to-cash-agents AI Weekly en 2026-10-07T14:01:47+00:00 Stuut Lands $52.5M Series B to Scale Order-to-Cash Agents https://aiweekly.co/alerts/huang-et-al-map-in-parameter-memory-methods-for-llms-on-two-axes AI Weekly en 2026-10-07T14:01:47+00:00 Huang et al. map in-parameter memory methods for LLMs on two axes https://aiweekly.co/alerts/finland-halts-work-at-googles-muhos-and-kajaani-data-centers AI Weekly en 2026-10-07T11:01:02+00:00 Finland Halts Work at Google's Muhos and Kajaani Data Centers https://aiweekly.co/alerts/common-sense-media-rates-chatgpt-for-teens-unacceptable-risk AI Weekly en 2026-10-07T11:01:02+00:00 Common Sense Media rates ChatGPT for Teens 'Unacceptable Risk' https://aiweekly.co/alerts/atlassian-commits-spend-to-openai-keeps-rovo-router-open AI Weekly en 2026-10-07T11:01:02+00:00 Atlassian Commits Spend to OpenAI, Keeps Rovo Router Open https://aiweekly.co/alerts/ctrlcache-cuts-interactive-video-world-model-compute-up-to-141x AI Weekly en 2026-10-07T11:01:02+00:00 CtrlCache cuts interactive video world model compute up to 1.41x https://aiweekly.co/alerts/nvidia-nemo-dcr-cuts-1t-rl-weight-sync-from-875-min-to-150-sec AI Weekly en 2026-10-07T09:40:32+00:00 Nvidia NeMo-DCR Cuts 1T RL Weight Sync From 87.5 Min to 150 Sec https://aiweekly.co/alerts/unreal-turns-a-frozen-llm-into-its-own-retriever-with-500k-params AI Weekly en 2026-10-07T09:40:32+00:00 UNREAL turns a frozen LLM into its own retriever with <500K params https://aiweekly.co/alerts/advsim2real-trains-4b-web-agent-against-adaptive-prompt-injection AI Weekly en 2026-10-07T09:40:32+00:00 AdvSim2Real Trains 4B Web Agent Against Adaptive Prompt Injection https://aiweekly.co/alerts/anthropic-pressed-vatican-to-soften-popes-ai-encyclical AI Weekly en 2026-10-07T09:12:01+00:00 Anthropic Pressed Vatican to Soften Pope's AI Encyclical https://aiweekly.co/alerts/ataraxos-beats-top-human-stratego-player-1500-the-compute AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats top human Stratego player, 1/500 the compute https://aiweekly.co/alerts/ataraxos-beats-top-stratego-player-on-8000-compute-budget AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats top Stratego player on $8,000 compute budget https://aiweekly.co/alerts/ataraxos-ai-beats-world-stratego-champion-15-1-4-in-nature AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos AI beats world Stratego champion 15-1-4 in Nature https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-at-1500th-the-compute-cost AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats Stratego champion at 1/500th the compute cost https://aiweekly.co/alerts/ataraxos-ai-beats-stratego-world-champion-in-nature-paper AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos AI beats Stratego world champion in Nature paper https://aiweekly.co/alerts/ataraxos-beats-top-stratego-player-15-1-for-under-8000 AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats top Stratego player 15-1 for under $8,000 https://aiweekly.co/alerts/ataraxos-defeats-strategos-top-player-15-1-4-in-nature AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos defeats Stratego's top player 15-1-4 in Nature https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-on-8000-training-budget AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats Stratego champion on $8,000 training budget https://aiweekly.co/alerts/ataraxos-beats-top-stratego-human-15-1-4-at-1500th-cost AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats top Stratego human 15-1-4 at 1/500th cost https://aiweekly.co/alerts/ataraxos-beats-strategos-top-pros-on-under-8000-compute AI Weekly en 2026-10-07T09:12:01+00:00 Ataraxos beats Stratego's top pros on under $8,000 compute https://aiweekly.co/alerts/google-arda-eye-screening-ai-crosses-one-million-patients AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA eye-screening AI crosses one million patients https://aiweekly.co/alerts/googles-arda-eye-ai-tops-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA eye AI tops one million patients screened https://aiweekly.co/alerts/google-arda-eye-screening-ai-crosses-one-million-patients AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA eye-screening AI crosses one million patients https://aiweekly.co/alerts/google-arda-eye-ai-crosses-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA eye AI crosses one million patients screened https://aiweekly.co/alerts/google-arda-eye-ai-tops-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA eye AI tops one million patients screened https://aiweekly.co/alerts/googles-arda-retinal-ai-passes-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA retinal AI passes one million patients screened https://aiweekly.co/alerts/googles-arda-retinal-ai-passes-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA retinal AI passes one million patients screened https://aiweekly.co/alerts/google-arda-eye-ai-tops-one-million-patients-across-three-countries AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA eye AI tops one million patients across three countries https://aiweekly.co/alerts/googles-arda-retinal-ai-crosses-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA retinal AI crosses one million patients screened https://aiweekly.co/alerts/googles-arda-retinal-ai-tops-1m-patients-in-three-countries AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA retinal AI tops 1M patients in three countries https://aiweekly.co/alerts/google-retinal-ai-tops-one-million-patients-in-three-countries AI Weekly en 2026-10-07T09:12:01+00:00 Google retinal AI tops one million patients in three countries https://aiweekly.co/alerts/googles-arda-retinal-ai-tops-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA retinal AI tops one million patients screened https://aiweekly.co/alerts/google-arda-tops-one-million-diabetic-eye-screens-globally AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA tops one million diabetic eye screens globally https://aiweekly.co/alerts/google-arda-retinal-ai-tops-one-million-patients-screened AI Weekly en 2026-10-07T09:12:01+00:00 Google ARDA retinal AI tops one million patients screened https://aiweekly.co/alerts/googles-arda-eye-screening-ai-tops-one-million-patients AI Weekly en 2026-10-07T09:12:01+00:00 Google's ARDA eye-screening AI tops one million patients https://aiweekly.co/alerts/boston-dynamics-names-ex-amazon-ai-chief-rohit-prasad-ceo AI Weekly en 2026-10-07T08:02:39+00:00 Boston Dynamics names ex-Amazon AI chief Rohit Prasad CEO https://aiweekly.co/alerts/openai-says-372-math-results-came-mostly-from-one-prompt AI Weekly en 2026-10-07T08:02:39+00:00 OpenAI Says 372 Math Results Came Mostly From One Prompt https://aiweekly.co/alerts/openai-launches-decisions-api-public-beta-on-gpt-6-luna AI Weekly en 2026-10-07T08:02:39+00:00 OpenAI launches Decisions API public beta on GPT-6 Luna https://aiweekly.co/alerts/googles-nano-banana-21-halves-image-prices-on-gemini-36-flash AI Weekly en 2026-10-07T08:02:39+00:00 Google's Nano Banana 2.1 Halves Image Prices on Gemini 3.6 Flash https://aiweekly.co/alerts/nathan-lambert-says-glm-53-release-produced-no-attack-wave AI Weekly en 2026-10-07T08:02:39+00:00 Nathan Lambert Says GLM-5.3 Release Produced No Attack Wave https://aiweekly.co/alerts/ex-ramp-engineers-melius-raises-25m-for-ai-ad-creative-agents AI Weekly en 2026-10-07T08:02:39+00:00 Ex-Ramp Engineers' Melius Raises $25M for AI Ad-Creative Agents https://aiweekly.co/alerts/alibaba-top-16-reverse-kl-matches-full-vocab-opd-on-3-pairs AI Weekly en 2026-10-07T08:02:39+00:00 Alibaba: Top-16 Reverse KL Matches Full-Vocab OPD on 3 Pairs https://aiweekly.co/alerts/alibaba-trace-matches-bf16-moe-rl-at-54x-fp4-rollout-speed AI Weekly en 2026-10-07T08:02:39+00:00 Alibaba TRACE Matches BF16 MoE RL at 5.4x FP4 Rollout Speed https://aiweekly.co/alerts/anthropic-lobbied-vatican-to-soften-popes-ai-consciousness-line AI Weekly en 2026-10-07T06:12:01+00:00 Anthropic lobbied Vatican to soften Pope's AI consciousness line https://aiweekly.co/alerts/google-ships-embeddinggemma-2-for-on-device-multimodal-retrieval AI Weekly en 2026-10-07T06:12:01+00:00 Google ships EmbeddingGemma 2 for on-device multimodal retrieval https://aiweekly.co/alerts/huatuogpt-3s-rl-only-recipe-tops-gpt-6-astra-on-healthbench AI Weekly en 2026-10-07T03:54:49+00:00 HuatuoGPT-3's RL-only recipe tops GPT-6 Astra on HealthBench https://aiweekly.co/alerts/nyu-llm-chain-of-thought-448759-faithful-across-tasks AI Weekly en 2026-10-07T03:54:49+00:00 NYU: LLM chain-of-thought 44.8–75.9% faithful across tasks https://aiweekly.co/alerts/duomatching-reports-80-human-preference-over-video-baselines AI Weekly en 2026-10-07T03:54:49+00:00 DuoMatching reports 80%+ human preference over video baselines https://aiweekly.co/alerts/openai-publishes-722-ai-generated-math-manuscripts-on-github AI Weekly en 2026-10-07T03:12:02+00:00 OpenAI publishes 722 AI-generated math manuscripts on GitHub https://aiweekly.co/alerts/willison-releases-llm-plugin-for-openais-decisions-api AI Weekly en 2026-10-07T03:12:02+00:00 Willison releases LLM plugin for OpenAI's Decisions API https://aiweekly.co/alerts/openai-publishes-722-ai-generated-math-preprints-on-github AI Weekly en 2026-10-07T03:12:02+00:00 OpenAI publishes 722 AI-generated math preprints on GitHub https://aiweekly.co/alerts/ard-joins-mcp-and-a2a-in-a-crowded-open-agent-protocol-field AI Weekly en 2026-10-07T03:12:02+00:00 ARD joins MCP and A2A in a crowded open-agent-protocol field https://aiweekly.co/alerts/anthropic-lobbied-vatican-over-popes-ai-consciousness-line AI Weekly en 2026-10-07T00:12:01+00:00 Anthropic lobbied Vatican over Pope's AI consciousness line https://aiweekly.co/alerts/wikimedia-says-rogue-openai-agents-probed-tools-flooded-servers AI Weekly en 2026-10-07T00:12:01+00:00 Wikimedia says 'rogue' OpenAI agents probed tools, flooded servers https://aiweekly.co/alerts/scholars-say-un-ai-panel-misdiagnosed-openaihugging-face-breach AI Weekly en 2026-10-07T00:12:01+00:00 Scholars Say UN AI Panel Misdiagnosed OpenAI–Hugging Face Breach https://aiweekly.co/alerts/photocraft-ships-rust-photoshop-clone-with-mcp-agent-server AI Weekly en 2026-10-07T00:12:01+00:00 PhotoCraft ships Rust Photoshop clone with MCP agent server https://aiweekly.co/alerts/apple-google-redraw-recording-for-always-on-ai-gadgets AI Weekly en 2026-10-06T21:12:01+00:00 Apple, Google redraw 'recording' for always-on AI gadgets https://aiweekly.co/alerts/anthropic-lobbied-vatican-to-soften-pope-leos-ai-encyclical AI Weekly en 2026-10-06T21:12:01+00:00 Anthropic lobbied Vatican to soften Pope Leo's AI encyclical https://aiweekly.co/alerts/openais-astra-proof-draws-fire-from-mathematician-andreas-thom AI Weekly en 2026-10-06T21:12:01+00:00 OpenAI's Astra proof draws fire from mathematician Andreas Thom https://aiweekly.co/alerts/peng-kleinberg-garg-large-saes-recover-features-in-order AI Weekly en 2026-10-06T21:12:01+00:00 Peng, Kleinberg, Garg: large SAEs recover features in order https://aiweekly.co/alerts/sciconbench-top-ai-research-agent-scores-just-0337-f1 AI Weekly en 2026-10-06T21:12:01+00:00 SciConBench: Top AI research agent scores just 0.337 F1 https://aiweekly.co/alerts/rothfeld-casts-anti-ai-resistance-as-a-humanist-movement AI Weekly en 2026-10-06T21:12:01+00:00 Rothfeld casts anti-AI resistance as a humanist movement https://aiweekly.co/alerts/utah-expands-health-ai-sandbox-names-six-outside-auditors AI Weekly en 2026-10-06T21:12:01+00:00 Utah expands health AI sandbox, names six outside auditors https://aiweekly.co/alerts/reddit-chatgpt-citations-drop-86-after-august-8-retrieval-tweak AI Weekly en 2026-10-06T21:12:01+00:00 Reddit ChatGPT citations drop 86% after August 8 retrieval tweak https://aiweekly.co/alerts/coxon-tells-stewart-ai-race-is-gambling-with-our-lives AI Weekly en 2026-10-06T21:12:01+00:00 Coxon Tells Stewart AI Race Is 'Gambling With Our Lives' https://aiweekly.co/alerts/anthropic-opens-three-tier-cvp-folding-in-project-glasswing AI Weekly en 2026-10-06T20:03:05+00:00 Anthropic Opens Three-Tier CVP Folding In Project Glasswing https://aiweekly.co/alerts/sierra-meta-publish-personal-agent-protocol-backed-by-walmart AI Weekly en 2026-10-06T20:03:05+00:00 Sierra, Meta Publish Personal Agent Protocol Backed by Walmart https://aiweekly.co/alerts/oppd-trains-qwen25-math-7b-to-beat-64-candidate-power-sampling AI Weekly en 2026-10-06T20:03:05+00:00 OPPD Trains Qwen2.5-Math-7B to Beat 64-Candidate Power Sampling https://aiweekly.co/alerts/sap-agrees-to-buy-techwolf-will-fold-skills-graph-into-joule AI Weekly en 2026-10-06T20:03:05+00:00 SAP Agrees to Buy TechWolf, Will Fold Skills Graph Into Joule https://aiweekly.co/alerts/linear-aligner-trained-in-minutes-boosts-tabpfn-3-and-tabfm AI Weekly en 2026-10-06T20:03:05+00:00 Linear Aligner Trained in Minutes Boosts TabPFN-3 and TabFM https://aiweekly.co/alerts/apo-paper-personalizes-llm-alignment-from-20-user-samples AI Weekly en 2026-10-06T20:03:05+00:00 APO Paper Personalizes LLM Alignment From 20 User Samples https://aiweekly.co/alerts/opentpu-ships-ai-developed-fpga-accelerator-for-qwen3-lfm25 AI Weekly en 2026-10-06T20:03:05+00:00 openTPU Ships AI-Developed FPGA Accelerator for Qwen3, LFM2.5 https://aiweekly.co/alerts/flai-raises-27m-for-ai-dealership-crm-booking-50000-a-month AI Weekly en 2026-10-06T18:38:26+00:00 Flai Raises $27M for AI Dealership CRM Booking 50,000 a Month https://aiweekly.co/alerts/anthropic-lobbied-vatican-to-soften-popes-ai-consciousness-line AI Weekly en 2026-10-06T18:12:01+00:00 Anthropic Lobbied Vatican to Soften Pope's AI Consciousness Line https://aiweekly.co/alerts/lambert-open-weights-cyber-risk-debate-misses-tradeoffs AI Weekly en 2026-10-06T18:12:01+00:00 Lambert: open-weights cyber risk debate misses tradeoffs https://aiweekly.co/alerts/alman-and-williams-crack-textbook-3sum-and-apsp-bounds AI Weekly en 2026-10-06T18:12:01+00:00 Alman and Williams crack textbook 3SUM and APSP bounds https://aiweekly.co/alerts/procedural-core-maps-1m-weights-to-vit-base-22pp-imagenet AI Weekly en 2026-10-06T18:12:01+00:00 Procedural Core maps 1M weights to ViT-Base, +2.2pp ImageNet https://aiweekly.co/alerts/chatgpt-gemini-become-ad-hoc-voter-guides-in-2026-midterms AI Weekly en 2026-10-06T18:12:01+00:00 ChatGPT, Gemini become ad hoc voter guides in 2026 midterms https://aiweekly.co/alerts/google-ships-embeddinggemma-2-740m-multimodal-embedder-apache-20 AI Weekly en 2026-10-06T16:58:56+00:00 Google ships EmbeddingGemma 2, 740M multimodal embedder, Apache 2.0 https://aiweekly.co/alerts/amazons-nova-25-sonic-voice-model-hits-bedrock-in-four-regions AI Weekly en 2026-10-06T16:58:56+00:00 Amazon's Nova 2.5 Sonic Voice Model Hits Bedrock in Four Regions https://aiweekly.co/alerts/a16z-top-1-of-us-ai-payers-drive-195-of-consumer-spend AI Weekly en 2026-10-06T16:58:56+00:00 a16z: Top 1% of US AI Payers Drive 19.5% of Consumer Spend https://aiweekly.co/alerts/antseed-launches-peer-to-peer-ai-inference-market-with-24m AI Weekly en 2026-10-06T16:58:56+00:00 Antseed launches peer-to-peer AI inference market with $2.4M https://aiweekly.co/alerts/anthropic-expands-claude-startups-with-45k-partner-stack AI Weekly en 2026-10-06T16:58:56+00:00 Anthropic Expands Claude Startups With $45K Partner Stack https://aiweekly.co/alerts/undobench-agents-complete-8354-of-tasks-recover-4672 AI Weekly en 2026-10-06T16:09:40+00:00 UndoBench: agents complete 83.54% of tasks, recover 46.72% https://aiweekly.co/alerts/split-learning-gradients-nearly-triple-document-leakage AI Weekly en 2026-10-06T16:09:40+00:00 Split-Learning Gradients Nearly Triple Document Leakage https://aiweekly.co/alerts/deskforge-trains-computer-use-agents-on-12m-desktop-scenes AI Weekly en 2026-10-06T16:09:40+00:00 DeskForge trains computer-use agents on 1.2M desktop scenes https://aiweekly.co/alerts/lift-loops-shared-dit-core-beats-dit-xl2-fid-by-334-points AI Weekly en 2026-10-06T16:09:40+00:00 LiFT loops shared DiT core, beats DiT-XL/2 FID by 3.34 points https://aiweekly.co/alerts/anduril-lands-29b-navy-contract-plans-37b-baltimore-shipyard AI Weekly en 2026-10-06T15:43:32+00:00 Anduril Lands $2.9B Navy Contract, Plans $3.7B Baltimore Shipyard https://aiweekly.co/alerts/asos-shares-fall-13-after-ransom-note-pushed-via-its-own-app AI Weekly en 2026-10-06T15:43:32+00:00 ASOS Shares Fall 13% After Ransom Note Pushed Via Its Own App https://aiweekly.co/alerts/technion-paper-adapts-frozen-tabpfn-to-tabular-anomaly-detection AI Weekly en 2026-10-06T15:43:32+00:00 Technion Paper Adapts Frozen TabPFN to Tabular Anomaly Detection https://aiweekly.co/alerts/anthropic-lobbied-popes-advisers-to-soften-ai-encyclical AI Weekly en 2026-10-06T15:12:01+00:00 Anthropic Lobbied Pope's Advisers to Soften AI Encyclical https://aiweekly.co/alerts/naver-labs-distills-transformer-memory-into-recurrent-agents AI Weekly en 2026-10-06T15:12:01+00:00 Naver Labs distills transformer memory into recurrent agents https://aiweekly.co/alerts/openais-kwon-apologises-in-sydney-for-medicare-hack-delay AI Weekly en 2026-10-06T15:12:01+00:00 OpenAI's Kwon apologises in Sydney for Medicare hack delay https://aiweekly.co/alerts/mistral-previews-1t-param-open-weight-large-4-le-chonk AI Weekly en 2026-10-06T13:58:40+00:00 Mistral Previews 1T-Param Open-Weight Large 4 'Le Chonk' https://aiweekly.co/alerts/kuaishous-kling-ai-taps-cicc-goldman-ubs-for-1b-hk-ipo AI Weekly en 2026-10-06T13:58:40+00:00 Kuaishou's Kling AI Taps CICC, Goldman, UBS for $1B+ HK IPO https://aiweekly.co/alerts/hla-wm-fixes-long-range-forgetting-in-video-world-models AI Weekly en 2026-10-06T13:58:40+00:00 HLA-WM Fixes Long-Range Forgetting in Video World Models https://aiweekly.co/alerts/flare-hits-97-of-cot-accuracy-on-gsm8k-at-39x-speedup AI Weekly en 2026-10-06T13:58:40+00:00 FLaRe Hits 97% of CoT Accuracy on GSM8K at 3.9x Speedup https://aiweekly.co/alerts/google-locks-36-gw-constellation-deal-890-mw-from-reactor-uprates AI Weekly en 2026-10-06T12:29:03+00:00 Google locks 3.6 GW Constellation deal, 890 MW from reactor uprates https://aiweekly.co/alerts/hadrian-raises-40m-series-b-for-agentic-ai-offensive-security AI Weekly en 2026-10-06T12:29:03+00:00 Hadrian Raises $40M Series B for Agentic AI Offensive Security https://aiweekly.co/alerts/signsplit-debuts-400m-seed-at-1b-to-pay-ai-data-contributors AI Weekly en 2026-10-06T12:29:03+00:00 SignSplit Debuts $400M Seed at $1B to Pay AI Data Contributors https://aiweekly.co/alerts/china-hosts-41-of-top-ai-talent-30-to-1-flow-favors-us AI Weekly en 2026-10-06T12:29:03+00:00 China Hosts 41% of Top AI Talent, 30-to-1 Flow Favors US https://aiweekly.co/alerts/google-arda-eye-ai-tops-one-million-patients-in-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google ARDA eye AI tops one million patients in three countries https://aiweekly.co/alerts/google-eye-ai-tops-one-million-patients-in-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google eye AI tops one million patients in three countries https://aiweekly.co/alerts/google-arda-eye-ai-tops-a-million-patients-in-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google ARDA eye AI tops a million patients in three countries https://aiweekly.co/alerts/googles-arda-retinal-ai-tops-one-million-patients-screened AI Weekly en 2026-10-06T12:12:01+00:00 Google's ARDA retinal AI tops one million patients screened https://aiweekly.co/alerts/googles-arda-retinal-ai-passes-1m-screens-in-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google's ARDA retinal AI passes 1M screens in three countries https://aiweekly.co/alerts/google-arda-eye-ai-tops-one-million-patients-in-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google ARDA eye AI tops one million patients in three countries https://aiweekly.co/alerts/googles-arda-eye-screening-ai-tops-one-million-patients AI Weekly en 2026-10-06T12:12:01+00:00 Google's ARDA eye-screening AI tops one million patients https://aiweekly.co/alerts/googles-arda-retinal-ai-crosses-one-million-patients AI Weekly en 2026-10-06T12:12:01+00:00 Google's ARDA retinal AI crosses one million patients https://aiweekly.co/alerts/google-arda-retinal-ai-crosses-one-million-patients-screened AI Weekly en 2026-10-06T12:12:01+00:00 Google ARDA retinal AI crosses one million patients screened https://aiweekly.co/alerts/google-arda-eye-ai-screens-one-million-across-three-countries AI Weekly en 2026-10-06T12:12:01+00:00 Google ARDA eye AI screens one million across three countries https://aiweekly.co/alerts/paper-tests-the-cardinal-sin-of-test-set-tuning AI Weekly en 2026-10-06T12:12:01+00:00 Paper tests the 'cardinal sin' of test-set tuning https://aiweekly.co/alerts/nature-health-31-us-cases-map-ais-wider-liability-web AI Weekly en 2026-10-06T12:12:01+00:00 Nature Health: 31 US cases map AI's wider liability web https://aiweekly.co/alerts/schiff-pitches-fda-for-ai-mandatory-training-data-retention AI Weekly en 2026-10-06T09:27:31+00:00 Schiff Pitches 'FDA for AI,' Mandatory Training-Data Retention https://aiweekly.co/alerts/fraunhofer-pflm-predicts-six-languages-from-synthetic-pretraining AI Weekly en 2026-10-06T09:27:31+00:00 Fraunhofer PFLM Predicts Six Languages From Synthetic Pretraining https://aiweekly.co/alerts/nature-breaks-down-the-science-behind-ai-extinction-fears AI Weekly en 2026-10-06T09:12:01+00:00 Nature breaks down the science behind AI extinction fears https://aiweekly.co/alerts/nature-explainer-ai-extinction-case-rests-on-untestable-claims AI Weekly en 2026-10-06T09:12:01+00:00 Nature explainer: AI extinction case rests on untestable claims https://aiweekly.co/alerts/anthropic-staff-warn-ai-could-kill-us-all-rand-calls-it-faith AI Weekly en 2026-10-06T09:12:01+00:00 Anthropic staff warn AI could kill us all; RAND calls it faith https://aiweekly.co/alerts/nature-probes-thin-science-behind-ai-extinction-warnings AI Weekly en 2026-10-06T09:12:01+00:00 Nature probes thin science behind AI extinction warnings https://aiweekly.co/alerts/anthropic-alignment-lead-pegs-ai-extinction-risk-above-10 AI Weekly en 2026-10-06T09:12:01+00:00 Anthropic alignment lead pegs AI extinction risk above 10% https://aiweekly.co/alerts/nature-dissects-the-ai-extinction-debate-after-coxon-exit AI Weekly en 2026-10-06T09:12:01+00:00 Nature dissects the AI extinction debate after Coxon exit https://aiweekly.co/alerts/anthropic-alignment-lead-puts-ai-extinction-risk-above-10 AI Weekly en 2026-10-06T09:12:01+00:00 Anthropic alignment lead puts AI extinction risk above 10% https://aiweekly.co/alerts/nature-probes-thin-science-behind-ai-extinction-warnings AI Weekly en 2026-10-06T09:12:01+00:00 Nature probes thin science behind AI extinction warnings https://aiweekly.co/alerts/nature-probes-thin-science-behind-ai-extinction-warnings AI Weekly en 2026-10-06T09:12:01+00:00 Nature probes thin science behind AI extinction warnings https://aiweekly.co/alerts/nature-audits-the-science-behind-ai-extinction-warnings AI Weekly en 2026-10-06T09:12:01+00:00 Nature audits the science behind AI extinction warnings https://aiweekly.co/alerts/rand-researcher-calls-ai-extinction-forecasts-more-like-faith AI Weekly en 2026-10-06T09:12:01+00:00 RAND researcher calls AI extinction forecasts 'more like faith' https://aiweekly.co/alerts/nature-scientists-push-back-on-anthropic-extinction-warnings AI Weekly en 2026-10-06T09:12:01+00:00 Nature: scientists push back on Anthropic extinction warnings https://aiweekly.co/alerts/khlaaf-in-nature-regulate-ai-labs-like-nuclear-and-aviation AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: regulate AI labs like nuclear and aviation https://aiweekly.co/alerts/ai-nows-khlaaf-tells-nature-ai-firms-cant-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 AI Now's Khlaaf tells Nature: AI firms can't self-regulate https://aiweekly.co/alerts/khlaaf-in-nature-ai-labs-cannot-keep-defining-ai-governance AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: AI labs cannot keep defining AI governance https://aiweekly.co/alerts/ai-nows-khlaaf-ai-labs-cannot-be-trusted-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 AI Now's Khlaaf: AI labs cannot be trusted to self-regulate https://aiweekly.co/alerts/ai-nows-khlaaf-ai-firms-cant-be-left-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 AI Now's Khlaaf: AI firms can't be left to self-regulate https://aiweekly.co/alerts/heidy-khlaaf-regulate-ai-like-nuclear-aviation-and-banks AI Weekly en 2026-10-06T09:12:01+00:00 Heidy Khlaaf: regulate AI like nuclear, aviation and banks https://aiweekly.co/alerts/khlaaf-urges-aviation-style-liability-for-ai-developers AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf urges aviation-style liability for AI developers https://aiweekly.co/alerts/khlaaf-in-nature-ai-labs-cannot-be-left-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: AI labs cannot be left to self-regulate https://aiweekly.co/alerts/ai-nows-khlaaf-ai-labs-cant-be-trusted-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 AI Now's Khlaaf: AI labs can't be trusted to self-regulate https://aiweekly.co/alerts/khlaaf-in-nature-ai-firms-cant-be-trusted-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: AI firms can't be trusted to self-regulate https://aiweekly.co/alerts/ai-nows-khlaaf-regulate-ai-labs-like-aviation-and-banking AI Weekly en 2026-10-06T09:12:01+00:00 AI Now's Khlaaf: regulate AI labs like aviation and banking https://aiweekly.co/alerts/khlaaf-in-nature-ai-firms-need-regulators-not-self-rule AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: AI firms need regulators, not self-rule https://aiweekly.co/alerts/khlaaf-ai-firms-need-aviation-style-oversight-not-self-rule AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf: AI firms need aviation-style oversight, not self-rule https://aiweekly.co/alerts/khlaaf-tells-nature-ai-labs-cant-be-left-to-self-regulate AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf tells Nature AI labs can't be left to self-regulate https://aiweekly.co/alerts/khlaaf-in-nature-ai-labs-need-aviation-style-oversight AI Weekly en 2026-10-06T09:12:01+00:00 Khlaaf in Nature: AI labs need aviation-style oversight https://aiweekly.co/alerts/ataraxos-beats-top-stratego-champion-for-under-8000 AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats top Stratego champion for under $8,000 https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-niemeijer-15-1-4-in-nature AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats Stratego champion Niemeijer 15-1-4 in Nature https://aiweekly.co/alerts/ataraxos-ai-tops-stratego-champion-at-1500-compute-cost AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos AI tops Stratego champion at 1/500 compute cost https://aiweekly.co/alerts/ataraxos-ai-beats-stratego-champion-15-1-for-under-8000 AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos AI beats Stratego champion 15-1 for under $8,000 https://aiweekly.co/alerts/ataraxos-defeats-top-human-stratego-player-15-1-in-nature AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos defeats top human Stratego player 15-1 in Nature https://aiweekly.co/alerts/ataraxos-ai-beats-stratego-world-champion-for-under-8000 AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos AI beats Stratego world champion for under $8,000 https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-niemeijer-15-1-4-in-nature AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats Stratego champion Niemeijer 15-1-4 in Nature https://aiweekly.co/alerts/ataraxos-beats-top-stratego-player-15-1-4-in-nature-paper AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats top Stratego player 15-1-4 in Nature paper https://aiweekly.co/alerts/ataraxos-defeats-top-stratego-champion-on-sub-8k-budget AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos defeats top Stratego champion on sub-$8K budget https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-on-8000-of-compute AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats Stratego champion on $8,000 of compute https://aiweekly.co/alerts/ataraxos-tops-stratego-champion-at-1500th-deepnash-compute AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos tops Stratego champion at 1/500th DeepNash compute https://aiweekly.co/alerts/ataraxos-beats-top-stratego-pro-at-1500-deepmind-compute AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats top Stratego pro at ~1/500 DeepMind compute https://aiweekly.co/alerts/ataraxos-ai-beats-world-stratego-champion-15-1-4-in-nature AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos AI beats world Stratego champion 15-1-4 in Nature https://aiweekly.co/alerts/ataraxos-beats-stratego-champion-using-1500th-the-compute AI Weekly en 2026-10-06T09:12:01+00:00 Ataraxos beats Stratego champion using 1/500th the compute https://aiweekly.co/alerts/google-arda-retinal-ai-tops-1m-patients-across-three-countries AI Weekly en 2026-10-06T09:12:01+00:00 Google ARDA retinal AI tops 1M patients across three countries https://aiweekly.co/alerts/google-arda-retinal-ai-passes-one-million-patients-screened AI Weekly en 2026-10-06T09:12:01+00:00 Google ARDA retinal AI passes one million patients screened https://aiweekly.co/alerts/googles-arda-eye-ai-crosses-one-million-patients-screened AI Weekly en 2026-10-06T09:12:01+00:00 Google's ARDA eye AI crosses one million patients screened https://aiweekly.co/alerts/yandex-mmd-post-training-cuts-diffusion-lm-perplexity-17-21 AI Weekly en 2026-10-06T07:56:31+00:00 Yandex MMD Post-Training Cuts Diffusion LM Perplexity 17-21% https://aiweekly.co/alerts/textreg-curbs-prompt-optimizer-overfit-beats-textgrad-by-118 AI Weekly en 2026-10-06T07:56:31+00:00 TextReg Curbs Prompt-Optimizer Overfit, Beats TextGrad by 11.8% https://aiweekly.co/alerts/omniconfess-scores-tokens-to-catch-omni-modal-hallucinations AI Weekly en 2026-10-06T07:56:31+00:00 OmniConfess Scores Tokens to Catch Omni-Modal Hallucinations https://aiweekly.co/issues/ai-got-too-expensive-so-companies-are-moving-to-cheaper AI Weekly en 2026-10-06T12:00:00+00:00 AI got too expensive, so companies are moving to cheaper models